Medical Error, Third Leading Cause Of Death? Even The Low Estimate Is Damning

Medical error is the third leading cause of death in America. Behind heart disease. Behind cancer. Ahead of strokes, accidents, and Alzheimer’s.

You have almost certainly seen that one. It is among the most shared health facts of the last ten years, and if it is true it is about the worst thing anybody could say about a hospital.

So we went and checked it. Here is the claim, here is the argument against it, and here is the part that matters most: even if the people arguing against it are completely right, what is left over is still bad enough to be angry about.

Where the claim comes from

In 2016 two surgeons at Johns Hopkins published a piece in the BMJ, one of the oldest medical journals in the world. The title was the claim itself: “Medical error-the third leading cause of death in the US” (Makary and Daniel, BMJ, 2016).

It travelled fast, and you can see why. Johns Hopkins, the BMJ, and a hard number on something plenty of people already suspected.

Here is the bit that gets left out. It was not a study. Nobody was examined. No records were reviewed. No new patients were involved at all. It was what journals call an analysis piece: the authors took figures from older research, averaged them together, and scaled the result up to the size of the country. That is a perfectly normal thing to publish. It is also a very different animal from a study, and the difference matters a great deal when a famous statistic gets built on top of it.

One thing we owe you. We are not printing the exact death figure that paper landed on, and we would rather tell you why than quietly leave it out. That paper has no summary published anywhere we can read, and neither does the main response to it. Both sit behind a paywall. The rule here is that we do not print a number we have not read in its own source, and we are not making an exception just because this one is famous and happens to suit our side. Every other number below is one we read ourselves.

The argument against it

The claim did not go unchallenged. Patient safety researchers at Toronto and Cambridge published a direct response in BMJ Quality & Safety that same year, and the title alone tells you the temperature: “Estimating deaths due to medical error: the ongoing controversy and why it matters” (Shojania and Dixon-Woods, BMJ Quality & Safety, 2016).

Rather than pick a side in that fight, we went to the people doing the slow work. Doctors who sit down with the records of someone who died in hospital and judge, case by case, whether that death had to happen.

One team collected every study of that kind they could find. Sixteen qualified, and eight of them, covering 12,503 deaths, were pooled together (Rodwin et al., Journal of General Internal Medicine, 2020).

3.1% of those hospital deaths were judged preventable (95% CI 2.2% to 4.1%). Scaled to the whole country, the authors put that at roughly 22,165 preventable deaths a year.

Then they asked a question most people skip: how many of those people would have lived much longer anyway? Among patients expected to live more than another three months, the preventable death rate dropped to between 0.5% and 1.0%, or about 7,150 a year.

The number of deaths due to medical error is lower than previously reported and the majority occur in patients with less than 3-month life expectancy. The vast majority of hospital deaths are due to underlying disease.

Rodwin et al., Journal of General Internal Medicine, 2020

England arrived at almost the same place by a different road. Trained reviewers read 100 randomly chosen deaths at each of 34 English hospital trusts and judged 3.6% avoidable (95% CI 3.0% to 4.3%). That rate fell over time, from 5.2% in 2009 to 3.0% by 2012/13 (Hogan et al., BMJ, 2015).

Two countries, two separate efforts, both landing near 3%. That is a long way below third leading cause of death, and we are not going to pretend otherwise.

Now hand the critics the whole argument

This is where most people stop reading, and it is where the thing gets interesting.

Give the critics everything they want. Say the famous claim is wrong. Use the lowest, most conservative, most carefully reviewed figure on the table.

You are still left with about 22,165 people a year who walked into a hospital and died of something that did not have to happen.

Now do the arithmetic the way you would for any other danger. Twenty-two thousand a year is about 426 people a week. A large passenger jet carries around 400. So the conservative, critics-win, checked-it-twice number is roughly a full jumbo jet falling out of the sky every week, all year, every year.

If that were actually happening, nobody would be bickering about whether it ranked third or eighth. Every plane in the country would be on the ground by Thursday.

And deaths are only the sharp end. The harm that does not kill anyone is far more common, and it has been measured far better.

Trained nurses screened a random sample of 2,809 hospital stays across 11 Massachusetts hospitals, with doctors checking every case they flagged (Bates et al., New England Journal of Medicine, 2023).

What they found	How often
Stays where at least one thing went wrong	23.6%, nearly one in four
Of those 978 events, ones that should not have happened	222 (22.7%)
Events serious enough to cause real harm	316 (32.3%)
Stays with a preventable harmful event	191 (6.8%)

The largest single category was medicine. Drug events made up 39.0% of everything that went wrong, ahead of surgery and procedures at 30.4%, nursing-care events like falls and bedsores at 15.0%, and infections caught in the hospital itself at 11.9%.

Nearly one stay in four. That figure is not an estimate, not an extrapolation, and not in dispute. Somebody counted.

So which number should you carry

Not the famous one. Not a shrug either.

The honest version is short enough to remember. Dying from a mistake in hospital is rarer than the viral statistic says. Being harmed in hospital is common. And the most common way it happens is medicine. All three are true at the same time.

There is one piece of genuinely good news, and we will give it because it is real. Looking only at what gets written on death certificates, deaths where medical treatment itself was the underlying cause fell 21.4% between 1990 and 2016, from 1.46 to 1.15 per 100,000 people (Sunshine et al., JAMA Network Open, 2019). Certificates undercount badly, because there is no box to tick for “the message never got passed on,” so the death gets filed under whichever organ gave out. But the direction is down, not up, and you would never guess that from how this gets talked about.

Why we are correcting a statistic that helps our own argument

We are openly sceptical of a trillion-dollar medical industry, and “third leading cause of death” is the best stick anybody in that fight has. We are putting it down anyway, for two reasons.

The first is that it is exactly what we demand of everyone else. We do not let a supplement company quote its best study and bury the rest, and we do not get a free pass because an exaggerated number happens to point our way.

The second is practical. A claim that falls apart under checking is a gift to the people you are arguing with. A hospital executive can take apart “third leading cause of death” in about sixty seconds, and while doing it, wave away the entire subject, including the one-in-four that is real and counted and happening this week. The fastest way to lose an argument is to overstate it.

What we are and are not saying

Claim	Our line
Medical error is the third leading cause of death	Not supported by studies that reviewed actual deaths
The claim came from a study	No. An analysis piece that reviewed no new patients
Preventable hospital deaths are rare	Rarer than claimed. About 3.1% of hospital deaths
Most of them are in people near the end of life	Yes. 0.5% to 1.0% among those expected to live beyond 3 months
The low estimate is reassuring	No. About 426 preventable deaths a week
Hospital harm is uncommon	No. Something went wrong in 23.6% of stays
Medicine is the biggest category of harm	Yes, 39.0% of events
Things are getting worse	Not on certificate data. The rate fell 21.4% from 1990 to 2016
You should avoid hospitals	Never. The alternative is usually far worse

Where the design shows up

There is an honest reason most hospital deaths could not have been prevented: the people in hospitals are, on the whole, extremely ill.

That is what a hospital is for. It gathers the sickest people into one building, which guarantees two things at once. Most deaths there will be deaths from disease. And any mistake made in that building lands on the person with the least strength left to survive it.

Holding both of those in your head at the same time is the whole skill. A system can be genuinely dangerous in places and still be the reason you are alive. Refusing that, in either direction, by insisting it is all a death machine or that the errors are overblown, costs real people real outcomes.

“A false balance is an abomination to the Lord, but a just weight is His delight.” Proverbs 11:1, NASB

The verse carries the conviction. Twelve thousand five hundred and three reviewed deaths carry the number.

What to do this week

The useful finding here is not the death toll. It is the one-in-four, and the biggest slice of that is medicine.

If you or someone you look after has a hospital stay coming, the single most valuable thing you can prepare is an accurate, current list of every medication and dose, written down and carried in. Drug events were 39% of everything that went wrong in that sample, and the most common reason a drug event happens in a hospital is that the team taking you in does not know exactly what you were already on.

We put the rest of what we would carry, and the evidence behind each line of it, in the card we would carry into a hospital stay.

Read this next: the card itself, and what is actually proven to help: nearly 1 in 4 hospital stays had an adverse event. How a side effect turns into a second prescription: the prescribing cascade. And how we grade a contested literature in general: two of medicine’s top editors say you can’t believe much of it.

Sources

1. Makary MA, Daniel M. “Medical error-the third leading cause of death in the US.” BMJ, 2016;353:i2139. https://pubmed.ncbi.nlm.nih.gov/27143499/
2. Shojania KG, Dixon-Woods M. “Estimating deaths due to medical error: the ongoing controversy and why it matters.” BMJ Quality & Safety, 2017;26(5):423-428. https://pubmed.ncbi.nlm.nih.gov/27733444/
3. Rodwin BA, Bilan VP, Merchant NB, et al. “Rate of Preventable Mortality in Hospitalized Patients: a Systematic Review and Meta-analysis.” Journal of General Internal Medicine, 2020;35(7):2099-2106. https://pubmed.ncbi.nlm.nih.gov/31965525/
4. Hogan H, Zipfel R, Neuburger J, Hutchings A, Darzi A, Black N. “Avoidability of hospital deaths and association with hospital-wide mortality ratios: retrospective case record review and regression analysis.” BMJ, 2015;351:h3239. https://pubmed.ncbi.nlm.nih.gov/26174149/
5. Sunshine JE, Meo N, Kassebaum NJ, Collison ML, Mokdad AH, Naghavi M. “Association of Adverse Effects of Medical Treatment With Mortality in the United States.” JAMA Network Open, 2019;2(1):e187041. https://pubmed.ncbi.nlm.nih.gov/30657530/
6. Bates DW, Levine DM, Salmasian H, et al. “The Safety of Inpatient Health Care.” The New England Journal of Medicine, 2023;388(2):142-153. https://pubmed.ncbi.nlm.nih.gov/36630622/

Article metadata retrieved from PubMed. Photograph from Pexels. Scripture quoted from the New American Standard Bible.

Educational content from True Health Solutions. We do not diagnose or prescribe, and nothing here is a substitute for care from someone who knows your history. The body heals. We write about what supports that design.